Reliability of a tool refers to the degree of consistency and accuracy with which it measures what it is intended to measure. If the evaluation gives more or less the same result every time it is used, such evaluation is said to be reliable. Consistency of a tool can be improved by limiting subjectivity of all kinds. Making items on the basis of pre-determined specific objectives, ensuring that the expected answers are definite and objective, providing clearly spelt-out scheme for scoring and conducting evaluation under identical and ideal condition will help in enhancing reliability.
Reliability is expressed numerically, usually as a coefficient. A high coefficient indicates high reliability and vice versa. If a test were perfectly reliable, the co-efficient will be 1. However, perfect reliability is impossible in practical situations.
Following are the most commonly used methods for assessing the reliability of a test.
1. The Test- Retest method
In this method the same test is re-administered shortly after the first administration and the two sets of scores are correlated to obtain the reliability of the test. The chief disadvantage of this method is that if the time interval between the two administrations of the test is short, the immediate memory affects, practice and confidence induced by familiarity with the test material may over estimate the reliability of the test. On the other hand, if the time interval is long, the real changes in terms of growth may under estimate the reliability of the test. Test re-test method is generally less useful than the other methods. The procedure for determining test-retest method is basically very simple:
a. Administer the test to an appropriate group.
b. After some time has passed, say a week, administer the same test to the same group.
c. Correlate the two sets of scores.
d. Evaluate the results.
2. The Equivalent/ Alternative/ Parallel Form Method
In this method two equivalent or parallel test are prepared and administered to the same group of subjects and the results in terms of two sets of the test scores are correlated to obtain the reliability of the test.
Here the two sets of the tests are identical in every respect such as they measure the same objectives, form the same content area, same number of items, same structure, same difficulty level and same directions for administration and scoring except the nature of the actual items.
Care should be taken to match test materials for content, objectives, difficulty and form and precautions must be taken not to have the items in the two forms too similar.
Even though it is too difficult to prepare two sets of test exactly in parallel in all its sense, it is one of the widely used methods to ensure the reliability.
Following procedure is adopted for the administration of this method:
a. Prepare two sets of test with same design and blue print.
b. Administer one form of the test to an appropriate group.
c. At the same session, or shortly thereafter, administer the second test to the same group.
d. Correlate the two sets of scores.
e. Evaluate the results.
3. The Split half method
In this method, the test is divided into two equivalent halves and it administered simultaneously to the same group. Then the scores of the half test are converted into the scores of the whole test by using Spearman- Brown Prophecy formula and then the two scores are correlated.
 |
Spearman –Brown Prophecy formula
The items of the test can be divided into two sets in a variety of ways. This method of reliability measures the internal reliability of the test and if the two halves do not correlate highly it suggest that they are not measuring the same thing.
The procedure for determining reliability under split-half method is as follows:
a. Split the test into two equivalent halves according to any approved or logical order
b. Administer the test simultaneously into the same group ( most common approach is to divide the test as odd and even)
c. Compute each subjects scores (two scores) separately
d. Correlate the two sets of scores.
e. Convert the coefficient of the half test into that of the whole test by using the formula.
f. Evaluate the result
No comments:
Post a Comment